sample_embedding_folder2/0744775.txt

Ticket Name: TDA2X / RTC 32.768 clock doesn't work

Query Text:
Other Parts Discussed in Thread: TPS65917-Q1, TDA2 Dear Mr, or Miss, I meet a problem , When I used PMIC (TPS659039Q1) PO9039A387IZWSRQ1+ TDA2SXBTQABCRQ1 (BGA760) design , the pin of TDA2X AE14,AD14 ( EXTERNAL 32.768 oscillator) is woriking , it has the RTC clock. The rtc clock can be detected when powered up even without firmware in TF CARD. But ,the next HW version we use PMIC(TPS65917) O917A133TRGZRQ1 instead of tps659039. These RTC CLOCK pins doesn't work , there is no 32.768 clock. please suggest what's the possible problem? By the way, I found the RTC_OSC_REG is related the settign , but I am not sure that the tp659039 can set the enable Initialization , but the tps65917 can't set Initialization.?? thanks in advance!

Responses:
Hi, Can you clarify which clock you are referring to? Is this the PMIC of processor clock? Here is our user guide that details how to connect the TPS65917-Q1 to the TDA2x processor: www.ti.com/.../slvuaj1 Thanks, Nastasha

We have connnected three cristal oscillator osc 20MHz， osc1 22.5792Mhz， RTC 32.768. We designed the first version based on 516582H_VAYU_EVM_13NOV2015_H。 The second version just used TPS65917 instead of TPS659039, modified based on the TPS65917 user guide power supply。 The difference is that RTC clck 32.768 doesn't work We wang to find out the reason, or how can I debug the problem.

Hi, Can you share the schematic of the PMIC? Thanks, Nastasha



Hi, How are you supplying VIO_IN? Also why is VRTC connected to SMPS4 (is this from the LDOVRTC?)? Thanks, Nastasha

Also, how are you turning on the PMIC? Are you using PWRON or POWERHOLD (GPIO_5)? Thanks, Nastasha

1) VIO IN is connected to EVM_3V3 2) VRTC_OUT IS CONNECTED TO smp4(vdds1.8) basd on TPS65917 usrer guide. it's not from LDOVRTC. 3) GPIO_5 is push up to the PS_EVM_3V3 to turn on the pmic.

Hi, Thanks for the clarification and diagrams. Can you share the full part number you are using (O917A1.....Q1)? Also, are you measuring at the SYNCCLKOUT pin? Thanks, Nastasha

Hi, The full part number of TPS65917 sample is 0917A133TRGZRQ1, and , I also measured the SYNCCLKOUT PIN ,that frequency is 2.135Mhz square wave, but the pin is not connected to the TDA2X , it is floated. any suggestion to where need to be tested or information from us? thanks and best regards Richard

Richard, Can you tell me where you are trying to measure the 32kHz clock? Thanks, Nastasha

We tested the oscillator of 32.768 , e.g. the pins AE14,AD14 of TDA2x. it doesn't have the wave. There is the schematic pictuire in above email. Thanks!

Hi, I don't see those pins being connected to the PMIC. I am looping in the processor team to help support. Thanks, Nastasha

Hi， Nastasha， Any update information about the RTC 32.768 clock problem？ thanks！

Hi, I have re-notified the processor team to look at this. The response may be delayed with people out for the holidays, but we will have someone look into this question further. Thanks, Nastasha

Richard, If there is no activity on the crystal connections, then likely there is a power issue. Have you confirmed all the voltage rails connected to SoC are powered and at correct voltage? Robert

Hi, If power supplies are good, then else you can check if it is due to Errata i856 32k Oscillator Fails to Start-Up at POR. Shortly, 32-k oscillator will not start without software intervention due to a bug: "Startup of 32k depends on user software writing to RTCSS GZ bit and only RTCSS can use the 32k clock (after writing GZ bit to 0), all other FUNC_32K_CLK targets (e.g. timers, mmc and gpio de-bounce) can only use SYSCLK1. Full features of RTC mode are still functional after the 32k software enables the oscillator." Regards, Stan

Hi @user4941012, We haven't heard back from you, I'm assuming you were able to resolve your issue. If not, just post a reply below (or create a new thread if the thread has locked due to time-out). Regards, Yordan

Hi Mr, I don't understand what's meaning. When we didn't insert TF card with firmware , Just power on , The first version HW can have the 32K clock internal with tps659039 pmic. Just ,the second version HW can't have with tps65919 pmic . If RTCSS GZ bit is writen by TP659039 OTP firmware?. If the tda2x ic has the bug , why the result is different between using TPS659039 an TPS65917? Sorry to ask again!

Hi, I'm not sure. Can you check the silicon revision of TDA2 on both designs? Are they the same? Also, I may be missed, but does 32-k clock run with TF card inserted (and SBL loaded)? Regards, Stan

Hi The sample versions of two version designs are the same. I am confirmed that the 32K-CLOCK runs up ,even no TF card inserted status. So , I am confused whether the TPS659039 and The TPS65917 have the different influence on TDA2X start. especilly on the 32K clock. thanks!

Richard, Nastasha, Any update on this item? Is this issue resolved? Thanks. Robert

Hi, Robert, This problem hasn't been resolved yet. I don't know which direction I can check ? Any suggestion ? and Did I express my problem clearly? Thanks!

As previously mentioned, the TDA2x errata states the oscillator fails to start-up after release of power-on-reset input. I understand the oscillator runs in one of your test configurations, and doesn't start in the other test configuration. However - it cannot be counted on operate in either configuration, as stated in the errata. There are differences between the two PMICs and how they are connected to the TDA2x device. Specifically, the 039 PMIC has dedicated power resources for the RTC domain, while the 917 PMIC does not. Thus - how the TDA2x RTC resources are connected between the two designs should be different. This might cause the different measure behavior. But in both configurations,the 32K oscillator errata still remains.